This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re:IPA Constant Propagation Optimization
- From: Razya Ladelsky <RAZYA at il dot ibm dot com>
- To: Steven Bosscher <stevenb at suse dot de>
- Cc: gcc at gcc dot gnu dot org, Mircea Namolaru <NAMOLARU at il dot ibm dot com>
- Date: Sun, 16 May 2004 12:56:10 +0300
- Subject: Re:IPA Constant Propagation Optimization
Hello,
> Does this also include function specialization/cloning if the function
is
> not always, but very often, invoked with the same constant argument?
Not for the first version. Possible in the future.
> Does the function tree have to be in some specific form, ie. do you
> expect every function tree to be still in SSA form? At what point
> in the compilation process would this pass operate? Right now we do
> not have any optimizers running before call graph optimizations, but
> I suppose you'll want to run a few optimizers (DOM, DCE,CCP) on each
> function tree before IPA.
For the IPA analysis, we need to have all the application's
methods in some intermediate representation at the same time.
For the moment this intermodule infrastructure exists only in parse
tree representation. IPA constant propagation will be performed on
all parse trees. Then the rest of compilation will be performed
one method at a time.
Razya and Mircea